霧島火山は，8世紀から多数の噴火記録があり，近年もきわめて活発な噴火を繰り返す日本有数の活火山である．歴史時代では御鉢，新燃岳，えびの高原-硫黄山の3火山において噴火が頻繁に発生し，複数の小火山が断続的に活動することも重要な特徴と言える．左図の『西嶽鹿倉絵図』は鹿児島藩都城島津家領内での鹿狩り域（鹿倉）を描いたものとされ，その上端に御鉢から立ち上る噴煙が見られる．本絵図は江戸時代後期に描かれたと推定されている（都城市／みやこのじょうデジタルアーカイブ^1））．本特集号では江戸時代後期に活躍した医師の橘 南谿が御鉢に登山した時（天明二年霜月，1782年12月）に遭遇した小噴火の記録を取り上げた（田島，2026）．絵図が橘の記録と同じ時期のものかは不明だが，江戸時代の御鉢の噴火を示す記録として貴重である．このほか，庄内地理志（巻54）には宝永七年八月下旬（1710年9月）に旗本の土屋（数馬）喬直を含む一行が幕府巡検使として都城から飫肥へ向かう道中での記録として「霧嶋山御鉢新火之儀御尋に付き」と記されている．この記録から1710年に近い時期に御鉢が噴火していたことを江戸幕府が把握していたと考えられる．右図は2018年の新燃岳噴火時に撮影され，火口内を溶岩で満した新燃岳とは対照的に静穏な状態が続く御鉢の様子が捉えられている．17世紀後半からの活動に限れば，御鉢が先行的に活動し，その後に新燃岳・えびの高原-硫黄山の活動が起こるパターンが4回繰り返されていることから，本写真は歴史上の霧島火山の活動を端的に切り取ったものかもしれない．

（西嶽鹿倉絵図（左）：都城島津邸所蔵，写真（右）：中田節也　2018年3月10日撮影；

説明：田島靖久）

注

1) https://adeac.jp/miyakonojo-city-shimazutei/catalog/ct00012889 [Cited 2026/3/22].

 Kirishima Volcano is a multiple volcano consisting of a complex of small volcanoes inside the volcanic edifice. For example, seven volcanoes developed as small volcanoes and produced many magmatic eruptions during the Holocene. Many eruption records of Kirishima Volcano, which indicate eruptions at the Ohachi, Shinmoedake, and Ebinokogen-Ioyama volcanoes, have been found in historical documents. Besides, eruptions at these volcanoes have been observed recently through geophysical monitoring and geological observations. In particular, a synchronous eruption occurred at the Shinmoedake and Ioyama volcanoes in 2018. Kirishima Volcano is a unique volcano, which has had multiple volcanic or eruption activities within its area in different time scales. Described are multiple eruption activities of similar eruption sequence cycles over different time scales exhibited by Kirishima Volcano, which are demonstrated by historical eruption records of historic eruptions, and material and geophysical analyses of recent eruptions. The current state of research and potential for future potential multiple volcanic activity are described.

 Research is reviewed on subsurface structures and the history of volcanic activity since the 20th century at Ebinokogen Iwo-yama volcano. The resistivity structure and hydrothermal system at Iwo-yama are revealed with electromagnetic, geodetic, and seismic surveys. The shallow low-resistivity layer distributed at around 200-700 m beneath Iwo-yama is considered to be a low-permeability clay-rich layer, above which there is a mixing zone of groundwater and volcanic fluids, and below which there is a reservoir of volcanic fluids observed as a pressure source of ground deformation. A general conceptual model in which a low-resistivity layer dominated by altered clay causes pressurization as a cap structure has been reported in relation to other hydrothermally active volcanoes, and the basic structure is understood to produce phreatic eruptions. Future work may include clarifying fluid supply systems from deep (> 4 km below sea level) to shallow (< 0 km below sea level) parts of the volcano and the internal structure of the shallow low-resistivity layer. Iwo-yama produced significant fumarolic activity from the early 20th century until the 1980s, but this activity declined from the mid-1990s and disappeared in 2007. A quiet period continued until 2013; then, shallow seismicity and ground deformation with seismic tremors began to be observed in 2014, suggesting the occurrence of fluid intrusion. Fumarolic activity resumed in December 2015. The fumarolic area expanded intermittently during the period 2016-2017, followed by a phreatic eruption in April 2018. Significant hydrothermal activity continued mainly at the Iwo-yama South Crater after the 2018 phreatic eruption, including repeated small eruptions in the period 2022-2023. Understanding the patterns and mechanisms of hydrothermal activity and transitions by continuing multi-parametric geophysical observations is essential for evaluating the potential for phreatic eruptions.

 Major element composition analyses are conducted on melt inclusions (MIs) and their host olivine phenocrysts in Takaharu scoria, which erupted at AD 1235 from Ohachi volcano, Kirishima volcano group, SW Japan, to investigate the magma plumbing system beneath the volcano. Olivine-hosted MIs are found in two of the three fall scoria units of the eruption: the first (ThT-a) and the second (ThT-b) units. In ThT-a scoria, olivine-hosted MIs are categorized into two clusters: one with S-rich basaltic compositions and the other with S-poor basaltic andesite compositions. In contrast, olivine phenocrysts from the ThT-b unit contained MIs of basaltic andesite composition. The compositions of the ThT-a S-poor and the ThT-b MIs are consistent with those of groundmass glasses of each scoria, respectively, indicating they record information on pre-eruptive conditions. The pre-eruptive H₂O content of the MIs is estimated based on olivine-melt CaO partitioning to be 6.6 ± 1.4 wt%. This corresponds to the depth of H₂O saturation in the melt of ∼17 km, which is consistent with H₂O-saturated plagioclase liquidus and is further supported by observed depths of deep low-frequency earthquakes. Based on the results, the magma plumbing system beneath the Ohachi volcano is discussed.

 Helium is present in very small amounts in volcanic gases, and its isotopic ratio (³He/⁴He ratio) sensitively reflects the state of magma underground. Spatial and temporal variations in the ³He/⁴He ratios of volcanic and hydrothermal fluids in and around the Kirishima Volcanic Group are discussed to clarify how magmatic fluids are incorporated into the shallow hydrothermal system and how their contributions reflect variations in magmatic activity. The spatial variation in the ³He/⁴He ratio of the fumaroles and hot/cold spring gases shows systematic ³He/⁴He decreases with increasing distance from the magma reservoir, the location of which is estimated as a pressure source of crustal deformation associated with the 2011 Shinmoe-dake eruption, to each sampling site. The spatial ³He/⁴He trend can be explained by a hydrodynamic dispersion model, in which crustal helium with a low ³He/⁴He ratio progressively leaches into an originally ³He-enriched magmatic fluid as it migrates through the local crust. However, the ³He/⁴He distribution is also controlled by the local permeable structure at shallow depths, resulting in a difference in the residence times of migrating fluids underground. The ³He/⁴He ratios of fumaroles in Iwo-yama slightly increase before Shinmoe-dake eruptions, and decrease after each eruption. This variation cannot be accounted for by the contribution of the radiogenic ⁴He relative to total helium in the fumaroles before and after the eruptions, because it is estimated that the amount of radiogenic ⁴He that volcanic gas can acquire during its migration from magma to the surface is negligible. Alternatively, the variation results from the change in mixing ratio of gases derived from two reservoirs having high and low ³He/⁴He ratios. Assuming that the magma chamber has a high ³He/⁴He ratio, the increase in the ³He/⁴He ratios of the fumaroles before the eruption would result from an increase in the supply of gas from the chamber to the Iwo-yama fumaroles. Once an eruption occurs at Shinmoe-dake, magmatic gas is effectively released through the volcanic vent, resulting in a decrease in its supply to the Iwo-yama fumaroles. Thus, the temporal variation of ³He/⁴He ratios in volcanic gases may reflect the pressure variation of the magma chamber.

 The SO₂/CO₂ ratio of fumarolic gas collected from vent (c) of Ebinokogen Ioyama volcano increased in response to a failed eruption in 2017 and a phreatic eruption in April 2018. Therefore, the SO₂/CO₂ ratio of fumarolic gas was thought to be a useful indicator for evaluating the volcanic activity of Ebinokogen Ioyama volcano. However, a gradual increase was observed in the ratio during the period 2020-2024, with no correlation to seismic activity. The SO₂/CO₂ ratio has not been an effective indicator of volcanic activity since 2020. The δ¹⁸O of fumarolic gas (h) increased during the period 2020-2024, reaching +4.4‰ in June 2023, although the temperature of fumarolic gas was 116°C. A fumarolic gas generation model is employed where a high-temperature magmatic vapor mixes with cold underground meteoric water, resulting in the formation of water vapor and thermal water. The fumarolic gas represents the water vapor. Applying the model, the above high δ¹⁸O (+4.4‰) requires a high mixing fraction of magmatic vapor. If the magmatic vapor temperature is assumed to be 900°C, no thermal water is generated and the water vapor temperature exceeds the observed fumarolic temperature. A possible range of magmatic vapor temperatures is 500 to 600°C, which is lower than the general magmatic temperature. Isenthalpic adiabatic expansion and conductive cooling as magmatic vapor ascends in the crust probably occurred after the degassing of magma. During the period from 2020 to 2024, the SO₂/CO₂ ratio of the two fumarolic gases (c and h) continued to rise with no correspondence with the number of earthquakes. The apparent equilibrium temperature (AET) calculated from the composition of the fumarolic gas also increased during the period from 2020 to 2024, suggesting a reduction in the sulfur sequestration effect for SO₂/CO₂ ratio and a reduction in the conductive cooling effect for AET. Effective indicators for evaluating volcanic activity at Ebinokogen Ioyama volcano include the δ¹⁸O of fumarolic gas (c), where the condensation effect of water vapor is observed in June 2024, and the AET of fumarolic gases (c and h). If the flux of magmatic vapor increases, the condensation effect of water vapor disappears, the δ¹⁸O of the fumarolic gas (c) increases, and the AET of the fumarolic gases (c and h) is also expected to increase, reflecting an increase in the temperature of the magmatic vapor.

 After the eruption of Ebinokogen Ioyama in April 2018, the formation of hot water pools at some craters was observed. Time-series changes in the chemical composition of the hot pool waters are studied over the period of more than six years from July 2018, with a view to monitoring time variations of water chemistry in the waning stage of geothermal activities. According to the observed time-series changes in fumarole and pool water temperatures, geothermal activities in the Ioyama fumarole area are in a waning stage in Period I (July, 2018–May, 2019) and Period II (Aug., 2019–July, 2021), whereas they show an increasing trend in Period III (Aug., 2021–Oct., 2022) and Period IV (Nov., 2022–Dec., 2024). The water chemistry of hot pools Y2a and Y2b in this area shows time-series changes from a rather high pH (up to pH = 2.2) and a low Cl/SO₄ molar ratio (< 0.02) in Period II to an extremely low pH (pH < 0) and a high Cl/SO₄ molar ratio (> 10) in Period IV, which is in accordance with that of geothermal activities. On the other hand, geothermal activities in the Ioyama-west fumarole area show a peak in Period II, when the temperature of the fumarole beside the hot pool W4 in this area reached 125°C. At around the same time, the hot pool water chemistry showed fluctuations in pH (pH = 0.5-1.6) and in Cl/SO₄ (Cl/SO₄ = 1.0-2.7). After hot pool W4 disappeared at the end of Period II, the water of another hot pool M8 in this area showed a rather constant chemistry, pH = 1.1-1.7 and Cl/SO₄ = 0.01-0.04. From observations in these two areas, time-series changes of pH in hot pool waters are likely to reflect a flux of the fumarole that supplies magmatic-derived Cl and SO₄ into the hot pool. On the other hand, Cl/SO₄ in hot pool waters could be affected by several factors. The observed low Cl/SO₄ in hot pools Y2a and Y2b in Period II is attributed to a process in an underground hydrothermal reservoir where HCl is separated from the fumarole due to preferential partitioning into the liquid phase. Whereas, the observed high Cl/SO₄ ratio in hot pools Y2a and Y2b in Period IV is likely to reflect the short residence time of magmatic gas in the hydrothermal reservoir due to an extremely high fumarole flux. It is difficult to find a good correlation between the observed time-series changes of hot water chemistry and magmatic activity of the Ebionokogen Ioyama represented by frequency of earthquakes. On the other hand, a sharp decrease of the Cl/SO₄ ratio in some hot pool waters is observed just after heavy rainfall in this area. This result suggests that the ground-water systems beneath the fumarole area are substantially affected by a large amount of meteoric waters being supplied. The contribution of meteoric water could be a specific factor, but it is also one of the important factors for understanding the geochemical process in the hydrothermal system of Ebionokogen Ioyama located in the area rich in rainfall.